Revision history [back]

your ANN needs "one-hot-encoded" responses to train, [1,0] for label 0, and [0,1] for label 1.

you can either change the csv file, so it has 16 data numbers and 2 response labels at the end:

0,0,1,0,0,0,1,0,0,0,1,0,0,0,1,0,0,1
1,1,1,1,1,0,0,1,1,0,0,1,,1,1,1,1,1,0

and use

Ptr<TrainData> datos = TrainData::loadFromCSV("/home/josejacomeb/QT/MLP_DosCapas/mlp_2capas.csv", 0, 16, 18);

(btw, not ROW_SAMPLE there !)

or, fix the responses after reading your unchanged csv:

 Mat responses(num_samples, 2, CV_32F, 0.0f);
 for (size_t i=0; i<num_samples; i++) {
      int id = (int)labels.at<float>(i);  // 0 or 1
      responses.at<float>(id) = 1;
 }

your ANN needs "one-hot-encoded" responses to train, [1,0] for label 0, and [0,1] for label 1.

you can either change the csv file, so it has 16 data numbers and 2 response labels at the end:

0,0,1,0,0,0,1,0,0,0,1,0,0,0,1,0,0,1
1,1,1,1,1,0,0,1,1,0,0,1,,1,1,1,1,1,0

and use

Ptr<TrainData> datos = TrainData::loadFromCSV("/home/josejacomeb/QT/MLP_DosCapas/mlp_2capas.csv", 0, 16, 18);

(btw, not ROW_SAMPLE there !)

or, fix the responses after reading your unchanged csv:

 Mat responses(num_samples, 2, CV_32F, 0.0f);
 for (size_t i=0; i<num_samples; i++) {
      int id = (int)labels.at<float>(i);  // 0 or 1
      responses.at<float>(id) = 1;
 }
 ann->train(datos->getTrainSamples(), 0, responses);

your ANN needs "one-hot-encoded" responses to train, [1,0] for label 0, and [0,1] for label 1.

you can either change the csv file, so it has 16 data numbers and 2 response labels at the end:

0,0,1,0,0,0,1,0,0,0,1,0,0,0,1,0,0,1
1,1,1,1,1,0,0,1,1,0,0,1,,1,1,1,1,1,0

and use

Ptr<TrainData> datos = TrainData::loadFromCSV("/home/josejacomeb/QT/MLP_DosCapas/mlp_2capas.csv", 0, 16, 18);

(btw, not ROW_SAMPLE there !)

or, fix the responses after reading your unchanged csv:

 Mat labels = datos->getTrainResponses(); 
 Mat responses(num_samples, 2, CV_32F, 0.0f);
 for (size_t i=0; i<num_samples; i++) {
      int id = (int)labels.at<float>(i);  // 0 or 1
      responses.at<float>(id) responses.at<float>(i, id) = 1;
 }
 ann->train(datos->getTrainSamples(), 0, responses);